SEQUENTIAL PENALTY HANDLING TECHNIQUES FOR SIZING DESIGN OF PIN-JOINTED STRUCTURES BY OBSERVER-TEACHER-LEARNER-BASED OPTIMIZATION

Despite comprehensive literature works on developing fitness-based optimization algorithms, their performance is yet challenged by constraint handling in various engineering tasks. The present study, concerns the widely-used external penalty technique for sizing design of pin-jointed structures. Observer-teacher-learner-based optimization is employed here since previously addressed by a number of investigators as a powerful meta-heuristic algorithm. Several cases of penalty handling techniques are offered and studied using either maximum or summation of constraint violations as well as their combinations. Consequently, the most successive sequence, is identified for the treated continuous and discrete structural examples. Such a dynamic constraint handling is an affordable generalized solution for structural sizing design by iterative population-based algorithms.

SEQUENTIAL PENALTY HANDLING TECHNIQUES FOR SIZING DESIGN OF PIN-JOINTED STRUCTURES BY OBSERVER-TEACHER-LEARNER-BASED OPTIMIZATION Keywords:

constraint handling , meta-heuristic algorithm , optimal structural design , sizing design , truss structures.